This invention relates to modified antibiotic peptides in particular for use in medicine. The invention relates also to compositions and methods for killing microorganisms, such as bacteria, viruses or fungi, and to methods for treating microbial infections.
The occurrence of serious bacterial and fungal infections is an increasing problem despite remarkable progress in antibiotic therapy. Each year there are more than 40 million hospital stays in the United States of America, and more than 2 million of those patients become infected in hospital. In 50-60% of these cases, antibiotic-resistant bacteria are involved. It is estimated that such diseases acquired in hospital lead to 60,000-70,000 deaths in the USA and up to 10,000 deaths in Germany.
This illustrates the necessity to continue the search for new antibiotics. Inducible antibacterial peptides represent a research area in which current biochemistry, immunology and active ingredient research come together. Peptide antibiotics, with a size of from 13 to more than a hundred amino acids, have been isolated from plants, animals and microbes (Boman, H. G. 1995 Annu. Rev. Immunol. 13: 61-92).
A single animal possesses about 6-10 antimicrobial peptides, each peptide often exhibiting a completely different activity spectrum (Barra, D. et al., 1998. FEBS Lett. 430: 130-134). It is known that the majority of antibacterial peptides, including the widely studied defensins, cecropins and magainins, act by a “lytic/ionic” mechanism. A permeabilising effect on the bacterial cytoplasm membrane is discussed as a common mechanism of action of these “lytic” peptides. A cationic, amphipathic structure, which forms hydrophilic ion (proton) channels in a lipid bilayer, is the basis for this activity. Through the occurrence of ions, the membrane potential necessary for many fundamental life processes is destroyed and the cell is thus killed. These lytic peptides often have a toxic effect on mammalian membranes at higher concentrations, which limits their suitability as possible medicaments. If proline is inserted into the sequence of the α-helical antimicrobial peptides, the ability of the peptides to permeabilise the cytoplasm membrane of E. coli falls, in dependence on the number of proline residues. From this point of view, it is amazing that some of the most active, natural antibacterial peptides, at least in relation to some gram-negative pathogens, belong to the family of the proline-rich peptides (Otvos, L. et al. 2000. Protein Sci. 9: 742-749).
The above-described side-effects could be overcome by antimicrobial peptides (AMP) which specifically recognise a bacterial protein or other intra- or extra-cellular components, without exhibiting cross-reactivity with mammalian analogs. This appears to apply to proline-rich antimicrobial peptides, including apidaecins, drosocin and pyrrhocoricin, which were originally isolated from insects. With the enormous variation in the size and biochemical properties, it is not surprising that the structure-action and conformation-action relationships are the focus of antibacterial peptide research. A complete study of the natural, antibacterial peptide repertoire for biological strength is not only important for general biochemical questions but is also of sustained interest for the pharmaceutical industry. Despite the problems of in vitro tests with peptide-based antibiotics, some natural, cationic antibacterial peptides have already reached the clinical trial phase (Boman, H. G. 1995 ebd.). While some of these peptides exhibited activity as topical (local) agents in the early clinical trial phase, others were active in systemic therapy. For example, the cationic protein rBPI 21, which is used for the parenteral treatment of meningococcemia, has completed the third phase of clinical testing (Boman, H. G. 1995 ebd.). The family of the proline-rich peptides (e.g. apidaecin, drosocin and pyrrhocoricin) kill bacteria not by permeabilisation of their membrane but bind stereospecifically to one or more target proteins. These possible interaction partners, the heat shock protein DnaK has hitherto been thoroughly researched (inter alia Boman, H. G. 1995), are inhibited by the proline-rich peptides and presumably the correct protein folding is inhibited, which ultimately leads to cell death. In addition, proline-rich peptides, in stark contrast to AMPs with a defined secondary structure such as melittin or gramicidin, do not appear in vitro to act either haemolytically or toxically on eukaryotic cells. In addition to the antimicrobial activity, the stability in mammalian serum (25%) especially has a decisive influence on the development of new peptide-based antibiotics. For example, drosocin is degraded within an hour, while pyrrhocoricin, with half-lives of 120 minutes, is considerably more stable to proteases. Presumably, not only are the N- and C-termini cleaved by amino and carboxy peptidases, but the peptides are also digested by endoproteases. Some of the metabolites formed thereby are stable to further decomposition but in most cases lose the antimicrobial activity (MIC values≧64 μg/ml).
In experiments conducted by Schneider M. and Dorn A. (2001. J lnvertebr Pathol. 78: 135-40), nymphs and pupae of the large milkweed bug Oncopeltus fasciatus from the family of the Lygaeidae were infected with two different gram-negative Pseudomonas species, and their immune response was analysed. While infection of the nymphs of O. fasciatus with the human pathogen Pseudomonas aeruginosa resulted in the death of all individuals after 48 hours, 71% of the individuals infected with the less pathogenic Pseudomonas putida survived for at least 96 hours. If the nymphs of the large milkweed bug were infected first with P. putida and after 24 hours with P. aeruginosa, the survival rate of the doubly infected individuals within the first 24 hours increased significantly to 73%. The probable induction of the synthesis of antibacterial peptides, by means of which insects defend themselves against incoming microorganisms within the context of their innate immune system, was then investigated. Four peptides (oncopeltus antibacterial peptide 1-4) were identified with molecular weights of 15, 8, 5 and 2 kDa and made responsible for the antibacterial activity. Sequence analysis according to Edman revealed, in addition to a partial sequence 34 amino acids long for peptide 1 (15 kDa), also the incomplete sequence of the proline-rich 2 kDa peptide 4. It was not possible clearly to identify the amino acids at positions 11 and the C-terminal sequence from position 19. The exact molecular weight is unknown.
A selection of hitherto known sequences of antibiotic peptides is listed in Table 1:
TABLE 1SEQPeptideSpeciesSequenceID No.Apidaecin 1aApisGNNRPVYI1melliferaPQPRPPHPRI Apidaecin 1bApisGNNRPVYI2melliferaPQPRPPHPRL DrosocinDrosophilaGKPRPYSP3melanogasterRPTSHPRPIRV Formaecin 1MyrmeciaGRPNPVNN4gulosaKPTPYPHL PyrrhocoricinPyrrhocorisVDKGSYLP5apterusRPTPPRPIYNRN-NH2 Metalnikowin 1PalomenaVDKPDYRP6prasinaRPRPPNM OncopeltusOncopeltusEVSLKGEG7antibacterialfasciatusGSNKGFIQpeptide 1GSGTKTLFQDDKTKLDGT OncopeltusOncopeltusVDKPPYLP8antibacterialfasciatusRP(X/P)Ppeptide 4PRRIYN(NR)
Apidaecin derivatives are disclosed in WO2009013262A1. Derivatives of oncopeltus antibacterial peptide 4 are disclosed in WO2010086401A1.
Different approaches at influencing the pharmacokinetic properties of pharmacological active ingredients are described in the literature. Organic polymers (such as e.g. polyethylene glycol) are also used thereby.
The use of polyethylene glycol (PEG) in pharmaceutical dosage forms for the controlled release of an active ingredient is known. A distinction must be made between two forms:                1. The introduction of the active ingredient into a crosslinked PEG hydrogel;        2. The direct bonding of a linear or branched PEG molecule to the active ingredient (known as PEGylation).        
The two forms differ not only in the pharmacokinetics but also especially in the possible administration routes. The hydrogel is administered locally. The PEGylated active ingredient is administered systemically (generally intravenously).
Hydrogels for the controlled release of an active ingredient are known inter alia from U.S. Pat. No. 7,291,673, US 2008/0014149 A1 and Yang J et al. 2010 (Macromol. Biosci., 10, 445-454).
The PEGylation of polypeptides is used in particular in order on the one hand to achieve controlled release over a desired period of time (retard effect) and on the other hand to delay the excretion of the active ingredient via the kidneys.
For the PEGylation of pharmaceutical active ingredients, reference may be made inter alia to U.S. Pat. No. 4,179,337, and also to the overview articles Kodera, M et al. 1998 (Prog Polym Sci. 23: 1233-71) and Veronese F M, Harris J M 2002 (Adv Drug Deliv Rev 54: 453-456).
Because an irreversible bonding of PEG impairs the pharmacological action of the active ingredient, a large number of known approaches for irreversible PEGylation exist.
Veronese F M (2001. Biomaterials 22: 405-17) describes in a general overview article the reversible PEGylation inter alia enzyme-catalysed to glutamine side chains by the enzyme transglutaminase. In an overview article by Roberts M J et al. 2002 (Adv Drug Deliv Rev 54: 459-476) too, several techniques for reversible PEGylation of peptides and proteins are mentioned, in particular hydrolysable ester bridges, reducible disulfide bridges.
EP1897561A, WO9930727, U.S. Pat. No. 6,180,095, U.S. Pat. No. 6,720,306, WO0243663 describe the use of a linker, which is cleaved by a 1,4- or 1,5-benzyl elimination. In several publications, this linker is combined in what is known as a double prodrug approach with an enzymatically cleavable group or a hydrolysable ester group (Greenwald R B et al. 1999. J. Med. Chem. 42: 3657-3667; Lee S et al. 2001 Bioconjug Chem 12, 163-169; Greenwald R B et al. 2003. Bioconjug Chem. 14(2): 395-403).
Another approach to reversible PEGylation is trimethyl lock lactonization (TML), which is disclosed inter alia in U.S. Pat. No. 5,965,119 and U.S. Pat. No. 6,303,569.
U.S. Pat. No. 7,585,837 describes an approach to reversible PEGylation by derivatisation of functional groups with 9-fluorenylmethoxycarbonyl (Fmoc) or 2-sulfo-9-fluorenylmethoxycarbonyl (FMS), which are readily cleaved by bases.
Guiotto et al. 2004 (J. Med. Chem. 47: 1280-9) describes the synthesis, characterisation and first in vivo tests of PEG conjugates of the antitumour agent 10-amino-7-ethylcamptothecin.
A protease-based prodrug strategy is based on peptide linkers which are bonded to conventional chemical active ingredients (“small molecules”) and contain cutting sites for proteases. The proteases that are used are mainly tissue-specific proteases or proteases which play a part in tissue remodelling (such as e.g. Cathepsin B, PSA (prostate specific antigen) and matrix metalloproteases (MMP), but also the serum proteases plasmin and urokinase. The field of application of the active ingredients is substantially cancer therapy. The protease cutting site is used for tumour targeting. In this connection, reference may be made to the overview article of Law B, Tung C H. 2009 (Bioconjug Chem. 20(9):1683-95). US 2004192769 mentions the approach of configuring the drug delivery system in such a manner that the peptide linker is cleaved not by serum proteases but only after being taken up into the target cell.
Li H et al. A 2010 (Angew. Chem. Int. Ed. 49, 4930-4933) apply a protease-based prodrug strategy to peptide active ingredients. Therapeutically active peptides are bonded via the peptide linker to an albumin binding domain (ABD). In the blood, this fusion peptide first binds to the serum protein albumin and is then cleaved by the protease thrombin (or human factor Xa).
There is a continued need for new antibiotics.
Desirable properties for peptide antibiotics are:                (i) an increased half-life in mammalian serum through a higher protease resistance and        (ii) unchanged or preferably increased antimicrobial activity against one or more bacterial strains, particularly human pathogens, or fungi or other microbial infections,        (iii) a reduced antigenic action and, as a result, a reduced immune reaction, and        (iv) the peptides are not toxic to human cells, including erythrocytes.        
The action of proline-rich antimicrobial peptides is very complex, because they must penetrate the cell membrane and pass into the cytoplasm in order to inhibit a specific intracellular bacterial target molecule, but without having a toxic effect on mammalian cells and blood cells. Another important point is the stability of the peptides or peptide derivatives to degradation by peptidases or proteases in blood and the bacteria. The ideal peptide therefore has a high antibacterial activity (low MIC values), no cell toxicity, no haemolytic activity and a half-life of several hours in blood.